The whole idea of a free service is to generate good will and potential customers. If you have a crappy free service, it generates bad will and discourages pote

The whole idea of a free service is to generate good will and potential customers. If you have a crappy free service, it generates bad will and discourages potential customers. So you are forced to maintain a decent free product, at least comparable to those of other people offering free products. An important thing to remember is that if you are offering a free service, your potential evaluation customers are likely to try out the free service before calling you, so if it is bad, you probably will not get the call. And, in general, maintaining decent quality is not free, although admittedly how expensive it is really depends on the details of the situation. I do agree that this analysis is wrong. It is common for companies trying to grow in market-share within a highly competitive end up losing money, if Box's problem was that it could only grow by losing money, it could just pick a "good enough" point and then scale back market share growth and instead concentrate on inside sales. I suspect that Box has other problems, and free business models often do have holes in them (people usually underestimate the costs of a free service that does not scare away customers), but there is far better analysis of free business models on the web than this article.